融合物理规律与数据驱动，提升气象预报长期稳定性。
Climate Physics Dynamic Matching
- 用变分框架结合平流物理先验与数据模型，不依赖模拟训练。
- 在小时级和月级预测上误差更低，且随时间累积更慢。
- 单块12GB消费级显卡即可训练，适合资源有限的研究者。
深度生成模型如流匹配和扩散模型在学习复杂动力系统方面展现出潜力，但通常作为黑箱处理，忽略底层物理结构；而基于偏微分方程的物理模型常因缺失源项或参数不确定性而不完整。本文提出气候物理动态匹配（ClimPhyDM），一种无需模拟的变分动力学信息框架，将平流型物理先验与数据驱动组件结合。该方法可捕捉未解析大气动力的随机性和多模态特征。在ERA5基准上，于小时级（42小时）和月级（5个月）分辨率下，ClimPhyDM优于ClimODE和GB-DM，保持更小误差并具有更好的时序稳定性，有效抵抗误差累积。其无模拟范式也使得仅需单块12 GB消费级显卡即可完成训练。
原文摘要 · Abstract (English)
Deep generative models such as flow matching and diffusion models have shown potential for learning complex dynamical systems, but typically act as black boxes that neglect underlying physical structure, while physics-based models governed by partial differential equations are often incomplete due to missing source terms, or uncertain parametrisations. We present Climate Physics Dynamic Matching (ClimPhyDM), a variational simulation-free dynamics informed framework for weather forecasting that combines an advection-type physics prior with data-driven components in a variational framework. % to capture the stochasticity and multi-modality of unresolved atmospheric dynamics. On the ERA5 benchmark at hourly (42-hour) and monthly (5-month) resolutions, ClimPhyDM outperforms ClimODE, and GB-DM, keeping the lower error at extended horizon, indicating improved temporal stability and resistance to error accumulation, while its simulation-free paradigm also enables training on a single modest 12 GB consumer GPU.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。